Background
A manufacturer wanted to automate a part-handling process currently done manually because the parts varied significantly in shape and orientation from batch to batch. Standard pick-and-place robotic systems, designed for uniform parts, could not reliably grip or orient the variable parts.
The Challenge
It was unclear whether any combination of gripper design, vision guidance, and motion planning could reliably handle the part variability at the required production speed, or whether the variability exceeded what automation could practically handle.
Technological Uncertainty
It was not known in advance whether a robotic system could achieve reliable grip and placement across the full range of part variation without excessive cycle time, or what gripper and vision configuration would be required.
Experimental Development
The team systematically tested combinations of adaptive gripper designs, vision-guided orientation detection, and motion planning strategies, measuring grip success rate, cycle time, and failure modes across representative part samples.
What Failed?
An initial rigid gripper with standard vision guidance achieved acceptable accuracy on regularly-shaped parts but failed to reliably grip parts at the extremes of the variation range, requiring redesign of the gripper mechanism itself.
Technological Advancement
The team developed a custom adaptive gripper combined with a refined vision-guided orientation algorithm that achieved reliable handling across the full part variation range, generating new technical knowledge about automating handling for this class of variable parts.
Potentially Relevant SR&ED Activities
- Systematic testing of gripper designs against variable part geometry
- Development and refinement of vision-guided orientation detection
- Cycle time and reliability benchmarking across iterations
What Would Generally Not Qualify
Running the finalized automated system in production on typical part batches would be routine operation and would not itself qualify as further eligible development.
Documentation
Grip success rate data across gripper iterations, vision algorithm development notes, and cycle time benchmarking records would support this claim.
